Skip to Content
Biotechnology and health

When Even Genome Sequencing Doesn’t Give a Diagnosis

The technology has been touted as a powerful diagnostic tool, but it doesn’t provide answers to everyone.
April 21, 2017
Sam D'Orazio

Four-year-old Beckett Edwards has had the works when it comes to genetic testing. And his family still doesn’t have an answer.

Soon after he was born, his parents, Eric and Tricia, noticed that his muscles were floppy. By age two and a half, Beckett had begun losing his 40- to 50-word vocabulary. Now he’s able to say only a handful of words and mostly babbles.

Like any parents whose child isn’t well, Eric and Tricia, who live in Los Angeles, want a diagnosis. Doctors suspect that Beckett has a genetic disorder and the answer lies in his DNA. Genetic tests often help physicians make diagnoses and, in some cases, steer patients to treatments (see “Slow Progress to Better Medicine”). That hope has taken the Edwards family on a diagnostic odyssey, through a series of increasingly detailed tests—ultimately sequencing Beckett’s complete genome. As the cost of gene sequencing drops, more comprehensive tests have become a realistic option for patients.

Beckett’s doctors first suggested standard genetic tests, including one for an inherited disorder called Prader-Willi, a rare disease that seemed to match some of his symptoms. But those tests came back negative.

With a handful of diseases ruled out, the family moved on to the next stage: whole-exome sequencing. That involves sequencing all of a person’s known genes—the “ex” in “exome” stands for “expressed,” meaning only the DNA whose job is to guide production of proteins. The exome represents only 1.5 percent of the six billion DNA letters in a person’s genome, but mutations in genes are most likely to cause disease. Jill Mokry, a genetic counselor at Baylor College of Medicine in Houston, calls the exome the “low-hanging fruit” of genetics.

All three Edwards family members received a whole-exome sequencing test. Since Beckett’s parents are healthy, any unusual DNA mutations they have could be ruled out as a cause of his illness. Exome sequencing was first introduced into specialized clinics about seven years ago. Studies have shown that it can render a diagnosis about 25 to 30 percent of the time. For example, a study published in Nature in March found that whole-exome sequencing successfully led to a diagnosis in 29 percent of 150 pediatric neurology patients, compared with 7 percent for standard genetic tests. Another study from February showed that a whole-exome test identified a genetic cause in 26.5 percent of 200 adults with inherited heart problems, compared with 18 percent for conventional genetic tests.

Eric and Tricia thought the exome would surely hold the genetic mutation responsible for Beckett’s symptoms. But when it revealed no obvious culprit either, a doctor at the University of California, Los Angeles, recommended that the family apply to the Undiagnosed Diseases Network, a study launched by the National Institutes of Health to help patients with previously unidentified medical conditions.

Beckett was accepted soon after the study opened in the fall of 2015, and last year he got a whole-genome sequencing test. Whole-genome sequencing provides a readout of all the DNA on a person’s chromosomes and has the potential to identify nearly all forms of genetic variation. Eric Edwards says that with this test, he definitely thought they’d get a diagnosis.

Beckett’s genome was sequenced by the HudsonAlpha Institute of Biotechnology in Alabama, where a study is under way to compare exome and genome sequencing. The institute’s price to sequence a whole genome is about $6,500, while sequencing an exome can run $5,000 to $5,500, says Liz Worthey, a geneticist at HudsonAlpha. Since the cost has become so similar for both tests, Worthey recommends that patients who haven’t received a diagnosis skip the exome and go straight to the genome.

The institute looked at the full genomes of 70 hard-to-crack cases who had previously received exome sequencing without getting answers. In 17 cases, or about 24 percent, the complete genome turned up a genetic mutation that was either the definitive or likely cause of the patient’s symptoms. In 47 others, or about 67 percent, patients’ genomes contained an unusual DNA variant that couldn’t be linked directly to their disease.

“Oftentimes in the genome we’re finding the variant, but that gene or that region is just not associated with disease as far as we know, so we’re not going to be able to say anything for sure,” Worthey says.

The analysis, presented last month at the Annual Clinical Genetics Meeting in Phoenix, showed that whole-genome sequencing can sometimes provide a diagnosis when previous whole-exome sequencing couldn’t. But it also reveals that the genome is still not well understood and that scientists don’t yet know all the alterations in the genome that may cause disease.

Isaac Kohane, of Harvard Medical School’s Department of Biomedical Informatics, says people analyzing a genome might simply miss a cause lying in plain sight—a so-called false negative.

“Our ability to distinguish what actually might cause disease is still quite crude,” Kohane says. “So we have an unknown false-negative rate.”

Another problem is that not all interpretations of the genome are equal: different research teams could analyze the same patient’s sequencing results differently. The sequencing itself is automated with a machine, and researchers use special software and online tools to sift through the large amounts of data produced (see “Internet of DNA”). Not all scientists are equally skilled at parsing genome files, though, Kohane says.

“Just as getting a biopsy on a lump in your breast merits a second opinion for a cancer diagnosis, so does rereading a genome,” Kohane says.

The Edwards family still has a few chances at a diagnosis. Matthew Herzog, a study coördinator for the Undiagnosed Diseases Network at UCLA, says researchers are exploring other ways to crack unsolved cases, such as RNA sequencing, which analyzes changes in how DNA is transcribed. It’s also possible to replicate suspect DNA changes in mice and zebra fish, to see if they cause symptoms similar to those seen in a human patient.

Herzog and his colleagues think they’ve identified a patient in New Zealand who has the same symptoms as Beckett. If they can get that patient’s genome sequenced, it may reveal a DNA mutation that both children share.

But Eric Edwards admits he and his wife are “at the point where we’re prepared that we may never know.”

Deep Dive

Biotechnology and health

Scientists are finding signals of long covid in blood. They could lead to new treatments.

Faults in a certain part of the immune system might be at the root of some long covid cases, new research suggests.

Gene editing had a banner year in 2023

This year, gene editing finally started living up to its potential

The first gene-editing treatment: 10 Breakthrough Technologies 2024

Sickle-cell disease is the first illness to be beaten by CRISPR, but the new treatment comes with an expected price tag of $2 to $3 million.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.